What is TechBio?
Why Cantos invests in engineering-led biology
At Cantos, we invest in the near frontier — companies that are pushing the boundaries of what is possible to advance human and environmental health. While there are many ways in which we can and must advance humanity, none is more obvious than through breakthrough discoveries within biology itself.
For nearly two decades, the tech revolution consumed nearly every industry except biotech. Biotech was too capital intensive, too conservative, and too time-consuming to be put on a trajectory similar to that of the software companies that defined the early 2000s — or so many thought. But while the funding structures and archetype of biotech companies remained shielded from change, the pace of innovation in the field picked up.
Enabled by advances in computation and automation, the field of biotechnology has made immense strides in the last two decades. We have sequenced the human genome, watched the costs of said sequencing drop from $2.7B to under $300, discovered precision engineering tools like CRISPR, developed entirely new treatment modalities such as CAR-T therapies, and have begun to trade scientist labor for machine intelligence to power drug discovery.
The very advancements in technology that have allowed for unprecedented biological innovation are changing the archetype of biotech companies themselves, enabling a new pace of biological innovation. Many refer to this new class of companies as “TechBio,” a term originally coined by Artis Ventures to reverse the emphasis and signal that the tech revolution has finally arrived in bio. Rather than biotech, this is the age of bio meets tech.
What is TechBio?
So what do TechBio companies look like? These companies are:
Engineering First: concerned more with using our understanding of biology to program bio and design with bio rather than to search for individual assets.
Founder-Led: started and operated by PhDs, engineers, and researchers often straight from university labs (which in turn has freed TechBio from the usual suspects: Boston-centric venture incubators.)
More Capital-Efficient: able to leverage advances in technology to de-risk and bring products/services to market quicker, with funding rounds similar to that of software companies.
Discovery Engines: focused on creating platform technologies that fuel the discovery and development of multiple asset classes.
Interdisciplinary: influenced by advances across multiple fields including computing, mechanical and electrical engineering, manufacturing, chemistry, and biology.
TechBio is the direct application of engineering to biology. While the portion of biology that we understand is still severely limited in scope, applying an engineering mindset to solving biological problems means we can still work with limited information. We see TechBio as being concerned with four main components:
- Reading bio — understanding genetic information (DNA, RNA, proteins, etc.) ex: next-gen sequencing
- Writing bio — synthesizing DNA, RNA, and protein constructs. ex: creating custom DNA oligonucleotides
- Programming bio — manipulating genetic information or small molecules for therapeutic purposes. ex: designing complex antibody drugs to treat metastatic cancer
- Delivering bio — routing biological information to the correct tissues and cells. ex: delivering RNA drugs to organ of interest
Think of these as the tools through which TechBio innovation can occur. Therapeutic innovation is primarily focused with programming bio. But in order to program a therapeutic, you have to be able to read bio to understand the molecular pathways and targets, write bio to design and test constructs, and deliver bio to get the drug to the right cells at the right time.
TechBio companies excel as platform-centric. Whereas traditional biotech is concerned with commercializing specific assets, TechBio focuses on powering the creation of multiple assets across indications. By being platform-centric, these companies are not defined by their asset pipeline but rather by their discovery engines. They can validate their platforms and generate early revenue by assisting critical R&D processes for other companies while simultaneously generating the data and expertise needed to create proprietary assets. They aren’t limited by specific scientific knowledge, instead relying on a combination of science, technology, and manufacturing to engage in partnerships and development. This allows for business model flexibility and provides a clear path to verticalization.
Platforms also unlock efficiencies. Take Moderna for example. Founded in 2010, Moderna has been developing and de-risking RNA technologies for the last decade. It was the steady refinement of their processes that allowed Moderna to develop a (now) approved COVID-vaccine in under two days. For an industry where the average time to clinic is 12 years, Moderna’s success is a resounding endorsement for TechBio.
What does TechBio look like?
Given that the end goal of TechBio is to create new medicines to treat (and prevent) disease, we’ll break up the TechBio stack into two categories of companies:
AI-enabled drug development (“AIDD”): computationally-powered drug discovery, optimization, repurposing, and delivery
Infrastructure: synthesis, manufacturing, scale-up, high-throughput sequencing, multi-omics, biotech SaaS, etc.
The former creates new medicines and is more focused with the programming aspect of TechBio, while the latter is essential in enabling the creation of such medicines — the read, write, and deliver parts of TechBio.
AI-enabled Drug Development
AI-enabled drug development is concerned with discovering, optimizing, and delivering new drugs. While most approved drugs to date have been small molecules, advances in bioengineering are pushing the trend towards next-generation medicines, including antibody conjugates, cell and gene therapies, and oligonucleotides (e.g. mRNA). With increasingly complex medicine surfaces the need for more extensive engineering strategies, which is where AI and automation shine.
Historically, new drugs have been discovered through a complex process that involves screening compound libraries against potential targets, performing experiments for promising candidates, and endless fine-tuning. It was a “find the needle in the haystack” problem. With machine learning and lab automation, we are now able to generate and sift through vast amounts of data, figuring out which drug/target pairs have higher chances for success by accounting for patterns only computers can find.
On the flip side, we can also program algorithms to discover new drugs with very little data as Nabla* is doing, circumventing the unknown complexities of biology. Rather than blindly searching for the needle, computational tools can either give us magnets or serve as the foundry to forge new needles. By applying AI, automation, and big data, we are able to move closer to the deliberate design of drugs. Early players in the AI-enabled drug discovery space include Recursion Pharma, Schrodinger, Atomwise, Insitro, Exscientia, and AbCellera, among others. Clinical readouts from these leaders will be instrumental in laying the foundations of AIDD as a field.
Beyond discovering and optimizing new drugs with AI, we can also use computational platforms to discover new delivery strategies for drugs. A major bottleneck of drug engineering is delivery. You can design a drug that acts perfectly on a disease pathway, but it is useless if it cannot get to the disease pathway of interest. Several companies are designing platforms that can rapidly customize delivery systems for different classes of therapeutics, selling use of the platform as a service — Sixfold Bioscience* and Dyno Therapeutics are good examples of this, respectively for RNA and gene therapies. These companies spend time learning how molecules interact within the body, mapping organ systems, and designing new constructs that can transport therapeutics to the right destination.
Until now, pharmaceutical companies would take care of the delivery component as part of the drug design process, typically using off-the-shelf delivery mechanisms like stable nanoparticles or conventional wisdom like using higher doses. But as therapies become more powerful, such methods are rarely sufficient. In some cases, they are outright dangerous — as we saw with Jesse Gelsinger’s case, which not only cost an 18 year old his life but also set the field of gene therapy back a decade. Instead, pharmaceutical companies can partner with delivery optimization startups for a customized solution based on molecules of interest, dosage, and delivery sites among other properties. Optimizing targeted delivery will lead to a decrease in off-target toxicities and an increase in therapeutic efficacy.
A major consideration for AIDD startups is whether to remain a horizontal platform, partnering with established pharma companies, or to vertically integrate into a full-stack drug development company themselves based on data generated from partnerships. This is not a binary choice, however. In our experience, many start horizontally integrated and move towards verticalization over time — a trend we have come to prefer.
Better treatments are the end-game for TechBio. We anticipate that there will be many successful companies in the AI-enabled drug development space, each differentiated by computational approach, point of action in the drug discovery pipeline, and asset classes enabled. Such companies may very well displace much of top 10 biopharma in the next two decades with their platform-oriented approaches.
The other side of TechBio lies in enabling the creation of medicines. These are the technologies permitting this new capital-efficient pace of drug development that we are seeing. Such technologies span everything from high-throughput sequencing to DNA synthesis to cell-therapy manufacturing. While we can’t cover all classes of infrastructure tech here, we’ll highlight some notable categories below.
High-throughput (HT) Technologies
Most innovation in bio requires generating and sifting through immense amounts of data to first understand the biology. Historically, this has required time-consuming (and often imprecise) assays and experiments. High-throughput screens employ automation to rapidly process experiments and mine biological pathways for insights. Such HT technologies are the basis for Illumina and new precision medicine companies. Beyond NGS, next generation high-throughput technologies unlock knowledge of cell dynamics down to the single-cell level. These technologies include microfluidics-based chips, multi-omics platforms, and advanced microscopy.
Operating Systems for Biotech
TechBio is largely enabled by big data. While wet-lab experiments and hardware will always remain a key part of bio companies, software and data are becoming increasingly common to gain a deeper understanding of biological interactions and patterns. Thus, there is a need for software solutions and lab automation that both help organize big data and accelerate R&D. Emerald Cloud Lab and Strateos are the north-star examples for what experimentation could look like — entirely virtual, streamlined, and (eventually) cost-efficient. Such a concept has yet to plant firm roots in industry given how complex it is to mechanize research, a stage where troubleshooting is arguably the most insightful step. Still, we expect “virtual CROs” are not that far out, with academia serving as an ideal place to pilot the concept — especially as we see an increasing number of graduate students looking towards commercializing their research.
Oligonucleotides, strands of genetic material like DNA and RNA, are the fundamental code of biology. There have been significant strides in genome sequencing and understanding this genetic code (reading), but there still remains a bottleneck in oligo synthesis (writing). Given that nearly all TechBio companies rely on creating customized DNA/RNA sequences for the development of platforms and/or therapeutics, there is a huge need for the rapid, customizable, and scalable production of oligonucleotide fragments. Companies innovating in this space include Twist Bioscience, Codex, Elegen, and Ansa Biotechnologies.
A therapeutic is no good if it doesn’t reach the bedside. Even after a new therapy gains FDA approval, there are still insurmountable challenges in getting it to patients. For cell and gene therapies, a major bottleneck lies in manufacturing — harvesting cells, modifying them, and scaling production in a timely manner. While there are several approved therapies, there is still a lack of scalable solutions for manufacturing since many of the therapies are individualized to each patient’s own cells — one of the reasons we see potential for allogeneic therapies. Companies like Cellares are creating lab-in-a-box type factories that bring manufacturing facilities closer to patients. Others are designing in vivo platforms to administer therapies without having to extract patient cells.
Also in the category of manufacturing is cell-free synthesis (CFPS) — producing proteins outside of living cells. CFPS lends scientists direct control over synthesis conditions, allows for scalable and quick production, and enables industrial applications since toxic proteins can be produced without consequence to living cells. Swiftscale Biologics and Arbor Bioscience are two players in this space.
Funding TechBio Companies
Funding structures are inevitably changing with this new archetype of biotech company. Rather than being incubated by large venture funds, TechBio companies often start fundraising from tech investors, bringing on larger biotech-focused funds in later rounds. This structure — which we emphatically believe is a superior model — lets founders retain more ownership and diversifies the approach to company building. For founders looking to raise early-stage capital, we have compiled a list of other tech funds that invest in TechBio. This is by no means a comprehensive list and we welcome any suggestions or edits.
The rate of biological innovation has already outpaced Moore’s law, but the TechBio revolution is only just getting started. New types of founders are building new types of biotech companies. Solutions are being uncovered to age-old biotech problems like protein folding and cancer detection. Possibly most exciting, multidisciplinary teams and brand new tools are being created to drive these biological insights. We believe the next decade will be instrumental for advancing human health.
You might argue that the role of tech in bio is not a new concept. While it is true that applying computation to rational drug design has been around for multiple decades, the power of TechBio as a whole is only just starting to materialize. We see now as the start of the “TechBio revolution” due to four major forces converging:
(i) increasingly advanced technology to discover biological insights
(ii) nimble company structures
(iii) deeper understanding of genomic architecture
(iv) heightened global interest in biological innovation
More and more life-science companies are being started by young, first-time founders who are often straight out of grad school. An increasing number of TechBio IPOs have demonstrated public interest in bio. And notably, there are more women founders in bio than in any other sector of tech. Though not without challenges, all this to say there is no doubt that the energy engulfing TechBio right now is special. We couldn’t be more excited to be a (small) part of these big stories as they are being written. If you are a founder engineering biology, drop us a line. We’d love to hear from you.
*Disclosure: Cantos is an investor in the companies marked with an asterisk.